EVALUATION OF HEAT AS A TRACER TO QUANTIFY LONGITUDINAL WATER FLOW IN THE HYPORHEIC ZONE

Quantifying flow in the hyporheic zone remains a significant challenge in stream hydrology. Naturally occurring variations in temperature have been used to quantify water flow in streambeds using vertical and transverse-vertical profiles, but not yet in a longitudinal-vertical profile. While general patterns exist for temperature signals in vertical profiles, longitudinal variations in heat are not necessarily as predictable. We collected vertical streambed temperature profiles at the upstream and downstream edges of a riffle in the lower American River, California. The upstream temperature profile shows significant diurnal fluctuation in the subsurface, while the downstream profile shows less diurnal fluctuation at all depths. Because these temperature profiles are distinct and were collected at the edges of a geomorphic feature, they should be useful as boundary conditions in a longitudinal heat transport model.

With these data, we conducted a series of numerical experiments using the USGS program VS2DH, to determine if longitudinal flow through the hyporheic zone can be reasonably inferred from shallow temperature profiles. Simulations performed with only longitudinal water flow through the streambed indicate that, when flow rates are varied over an order of magnitude, distinct temperature signals result at observation points in the upper 60 cm of the domain and 19 m and 38 m from the upstream boundary. For flow rates of 10^-4 cm/s and higher, three general qualities are present when comparing temperature signals: higher flow rates result in higher overall temperature (during summertime conditions), higher amplitude of diurnal fluctuation, and quicker response to day-to-day temperature change. These qualities imply that longitudinal flow inferred from model calibration to temperature may be useful for order-of-magnitude estimation of flow rate. Temperature signals resulting from different flow rates within an order of magnitude are similar, as are temperature signals resulting from flow rates below 10^-4 cm/s. Thus, quantifying longitudinal water flow by tracing natural heat variations may be limited to order-of-magnitude estimation of flow rates above 10^-4 cm/s.